kras99 - Fotolia


Office 365 AI gets practical for workers

Artificial intelligence used to be the province of academia. But today, flavors of it, like Microsoft AI, are having a real-world impact on work.

Lately, artificial intelligence has gotten a good deal of attention. That's partly because business can benefit...

from these systems by automating previously time-consuming but low-level tasks. AI can also help search for and retrieve information that is relevant to workers' daily tasks. Once used largely in academia and plot projects, artificial intelligence has become more central to workers' jobs by making business processes more efficient and "intelligent."

As a result, various providers are now moving to build artificial intelligence into their platforms by having computers observe workers' behavior in applications and intelligently "learn" and take action based on this user behavior. Artificial intelligence and machine learning are at the core of Office 365 services, Microsoft's productivity and collaboration suite. Microsoft's Office 365 AI moves have included Office Graph, Delve and other aspects of its services. The idea is to use machine learning to automate certain tasks, serve up relevant content and ideas, and, ultimately, enhance human work.

Unlike its competitors Google and, to some extent, IBM, Microsoft has tried to apply machine learning advances in "practical" ways that immediately benefit customers. Over the past few years, Microsoft has begun integrating machine learning technologies into email in Outlook (and Exchange), collaboration in SharePoint, and analytics with PowerBI and Cortana Analytics.

Bringing AI to these applications can mean enhanced productivity for users.

Office apps and Microsoft AI

Microsoft has incorporated machine learning into various elements of the suite. In particular, a feature called Tell Me now appears in the menu bar, which uses machine learning to divine answers to workers' Office feature questions. Type in a question about a function, task, or feature and Tell Me almost immediately begins to build a menu list of potential answers. Tell Me can execute document changes, such as changing text color, by showing a user the steps and simply requesting a confirmation.

In Outlook, machine learning removes clutter from and organizes a workers' inbox. There are two features here -- "clutter" and "focused inbox." Focused inbox refines the clutter feature by monitoring how users treat mail items: as they work with mail in their inbox, this feature learns which mail is important and the messages a user might ignore or delay opening. Outlook automatically begins to create two distinct views -- focused and everything else. The focused view includes email messages that, based on your inbox behavior, are important to you. The "other" view includes less-important email messages.

Office 365 AI and analytics

Microsoft's analytics applications, PowerBI and Cortana Analytics, are two direct machine learning beneficiaries. PowerBI brings analytics to a wider array of users within a company. Its machine learning features, like Office, are subtly applied. First, when you import data from Excel or nonrelational sources, PowerBI can infer relationships and automatically construct data views. This allows nontechnical users to quickly load data into a data model and immediately begin developing reports or an analytics dashboard. In addition, PowerBI uses natural language to describe data queries. Combined with the inferred data relationships, PowerBI can construct relatively complex data views without more skill than asking the right question.

Though not officially part of Office 365, one of the most obvious machine learning applications is in the Cortana Intelligence Suite. Building on Microsoft's digital assistant, Cortana, Microsoft has combined Azure Machine Learning capabilities with its data management tools. This has produced a business intelligence (BI) suite that includes more natural human interaction (speech and natural language queries) and predictive analysis, as well as prescriptive capabilities. In other words, in the same way that Cortana is the digital assistant in Windows capable of telling you what's on your schedule or finding a reservation at a local restaurant, Cortana Analytics uses similar capabilities with enterprise BI services. This combination enables more friendly interaction with BI tools, as well as the ability to automate some enterprise tasks based on your data.

Chatbots and automated self-service

Recently, Microsoft released Skype "bots." This feature enables humans to interact with chatbots that can assist with quick actions or tasks. For example, you can use the Hipmunk bot to assist in booking travel or buy tickets to a concert with StubHub's chatbot. This service, too, is an extension of the same capabilities used in Cortana and Cortana Analytics. But it's applied as a less-structured way and can be integrated into a basic communication tool.

Chatbots are compelling opportunities for the enterprise. As Microsoft continues to merge the consumer and business versions of Skype, these capabilities become available to firms. Microsoft has already released its bot framework. The framework allows developers to construct their own bots, using Azure machine learning. Consumer scenarios involving travel or product purchase can quickly translate to real-time inventory queries or summaries of a recent employee meeting.

Microsoft's AI journey

In much the same way that SharePoint commoditized document management and Excel/BI tools did the same for data, Microsoft now has a vision to democratize BI. Microsoft has begun to apply machine learning and, more broadly, artificial intelligence to a range of practice concerns. As Microsoft CEO Satya Nadella stated at Ignite conference, "[They're] not out to play games."

In real ways, Microsoft has put machine learning to work to improve product experiences. Whether it's creating a PC-based digital assistant in Cortana or improving help functions, Microsoft's Office 365 AI approach continues to build services that help workers with everyday tasks.

For more, check out all our Microsoft Ignite coverage here.

Next Steps

Microsoft AI pervades the stack, from Office 365 to Azure

What to expect at Microsoft Ignite

SharePoint mobile apps get a thumbs-up

Dig Deeper on Text analytics and natural language processing software